https://communities.sas.com/t5/Statistical-Procedures/GLIMMIX-Specifying-a-covariance-matrix/m-p/39300#M1650 Hello ,<BR /> <BR /> I have a problem with the proc GLIMMIX . I use it for fit a mixed model Y=XB + Zv +e where XB is the fixed effects and Zv is the random effect . <BR /> The vectors v and e follow a Gaussian law . the covariance matrix of v is G and the covariance matrix of e is R .<BR /> <BR /> I want to fix the matrix R than i know. <BR /> I know i can use " PARMS " with the option "hold=" but my matrix R is 300x300. <BR /> I would like set R from a data. <BR /> Can you help me to find a solution ?<BR /> <BR /> Thanks!<BR /> <BR /> (Sorry for my english... ) Wed, 23 Jun 2010 08:46:53 GMT deleted_user 2010-06-23T08:46:53Z https://communities.sas.com/t5/Statistical-Procedures/GLIMMIX-Specifying-a-covariance-matrix/m-p/39300#M1650 Hello ,<BR /> <BR /> I have a problem with the proc GLIMMIX . I use it for fit a mixed model Y=XB + Zv +e where XB is the fixed effects and Zv is the random effect . <BR /> The vectors v and e follow a Gaussian law . the covariance matrix of v is G and the covariance matrix of e is R .<BR /> <BR /> I want to fix the matrix R than i know. <BR /> I know i can use " PARMS " with the option "hold=" but my matrix R is 300x300. <BR /> I would like set R from a data. <BR /> Can you help me to find a solution ?<BR /> <BR /> Thanks!<BR /> <BR /> (Sorry for my english... ) Wed, 23 Jun 2010 08:46:53 GMT https://communities.sas.com/t5/Statistical-Procedures/GLIMMIX-Specifying-a-covariance-matrix/m-p/39300#M1650 deleted_user 2010-06-23T08:46:53Z https://communities.sas.com/t5/Statistical-Procedures/GLIMMIX-Specifying-a-covariance-matrix/m-p/39301#M1651 Check out the PARMSDATA= option. This should allow you to load in the values you want.<BR /> <BR /> Steve Denham Wed, 23 Jun 2010 12:34:06 GMT https://communities.sas.com/t5/Statistical-Procedures/GLIMMIX-Specifying-a-covariance-matrix/m-p/39301#M1651 SteveDenham 2010-06-23T12:34:06Z